Download The Origins of the Matrilineal Principle in Rabbinic Law

The Origins of the Matrilineal Principle in Rabbinic Law
Author(s): Shaye J. D. Cohen
Source: AJS Review, Vol. 10, No. 1 (Spring, 1985), pp. 19-53
Published by: Cambridge University Press on behalf of the Association for Jewish Studies
Stable URL: http://www.jstor.org/stable/1486271 .
Accessed: 19/01/2015 11:38
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
.
Cambridge University Press and Association for Jewish Studies are collaborating with JSTOR to digitize,
preserve and extend access to AJS Review.
http://www.jstor.org
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL
PRINCIPLE
IN RABBINIC LAW
by
SHAYE J. D. COHEN
According to rabbinic law, from the second century to the present, the
offspring of a gentile mother and a Jewish father is a gentile, while the
offspring of a Jewish mother and a gentile father is a Jew (albeit, according
to the Mishnah, a mamzer, a Jew of impaired status).' Each of these two
Research on this paper was supported by a generous grant from the Nisson Touroff Fund of
the Jewish Theological Seminary. References to the Mishnah, Tosefta, Babylonian Talmud,
and Palestinian Talmud are indicated respectively by M., T., B., and P. followed by the name of
the tractate. Unless otherwise noted, all translations in this essay are mine. This paper has benefited from the suggestions and advice of Mr. David Cherry, Mr. Leonard Gordon, and Professor Sarah Pomeroy. A capsule version of this essay has appeared in Judaism 34, 1 (Winter
1985): 5-13.
1. A mamzer is a male or female Jew (the feminine form of the noun is mamzeret) who,
because of the circumstances of his or her birth, may not marry a native-born Jew; if he or she
does, the children are mamzerim. I leave mamzer untranslated because the English terms "illegitimate" and "bastard" derive from a completely different legal system and do not accurately
reflect the meaning of the Hebrew.
19
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
20
SHAYE J. D. COHEN
rulings has its own history, as I shall show below, but it is convenient to
group them together under the general heading of the "matrilineal principle." Anthropologists and sociologists use the term matrilineal to describe
societies in which kinship is determined through the females and not the
males. Such societies once existed in ancient Egypt and Mesopotamia, and
can still be found in parts of Africa, India, and Polynesia.2 Although rabbinic society and family law have not yet been studied in the light of modern
anthropological and sociological theories, it seems clear that the kinship
patterns which characterize matrilineal societies are thoroughly foreign to
rabbinic society. With only a few exceptions, rabbinic family law is patrilineal. Status, kinship, and succession are determined through the father.
("The family of the father is considered family, the family of the mother is
not considered family," B. Bava Batra 109b.) Why, then, did the rabbis
adopt a matrilineal principle for the determination of the status of the
offspring of mixed marriages?
In the first section of this essay I examine the Bible and the texts of the
Second Temple period in order to determine the earliest attestation of the
matrilineal principle in Jewish sources. The second section is dedicated to
the Mishnah, the Tosefta, and the two Talmudim. In the third section I
assess various explanations for the origins of the matrilineal principle, and
in the fourth I attempt to reach a conclusion.3
2. Paul Koschaker, "Fratriarchat, Hausgemeinschaft und Mutterrecht in Keilschriftrechten," Zeitschrift fiir Assyrologie 41 n.s. 7 (1933): 1-89; David M. Schneider, Matrilineal
Kinship (Berkeley: University of California Press, 1961); in general see Jack Goody, ed.,
Kinship:Selected Readings (Baltimore: Penguin, 1971) and Jack Goody, ComparativeStudies in
Kinship (Stanford: Stanford University Press, 1969), chap. 3. For recent kinship studies see
Robert A. Oden, Jr., "Jacob as Father, Husband, and Nephew: Kinship Studies and the
Patriarchal Narratives," Journal of Biblical Literature 102 (1983): 189-205.
3. This article is the fourth in a series about conversion and intermarriage in antiquity; see
also "Conversion to Judaism in Historical Perspective: From Biblical Israel to Postbiblical
Judaism," Conservative Judaism 36, no. 4 (Summer 1983): 31-45; "From the Bible to the
Talmud: the Prohibition of Intermarriage," Hebrew Annual Review 7(1983): 23-39; "Was
Timothy Jewish (Acts 16:1-3)? Patristic Exegesis, Matrilineal Descent, and Rabbinic Law,"
Journal of Biblical Literature (in press). The most important scholarly contributions to the
study of the matrilineal principle are: Victor Aptowitzer, "Spuren des Matriarchats im
juedischen Schrifttum," Hebrew Union College Annual 4 (1925): 207-240 and 5 (1926):
261-297; Louis M. Epstein, Marriage Laws in the Bible and the Talmud(Cambridge, Mass.:
Harvard University Press, 1942), esp. pp. 174 and 194-197; Boaz Cohen, "Some Remarks on
the Law of Persons in Jewish and Roman Jurisprudence," Proceedings of the American
Academy for Jewish Research 16 (1946-47): 1-37, esp. 12-24; and Lawrence Schiffman, "At
the Crossroads: Tannaitic Perspectives on the Jewish-Christian Schism," in Jewish and
Christian Self-Definition, vol. 2, Aspects of Judaism in the Greco-Roman Period, ed. E. P.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
21
I
TheBiblicalPeriod
The preexilicportions of the Hebrew Bible are not familiarwith the
matrilinealprinciple.NumerousIsraeliteheroesand kings marriedforeign
women; for example, Judah marrieda Canaanite,Joseph an Egyptian,
Moses a Midianite and an Ethiopian, David a Philistine, and Solomon
women of every description.4Although Exod. 34:16 and Deut. 7:1-3
prohibit intermarriageonly with the Canaanites,a prohibitionthat was
supposedto have originatedwith the patriarchsAbraham(Gen. 24:3) and
Isaac(Gen. 27:46-28:1),some Israelitesextendedthe prohibitionto include
all foreigners(Judg. 14:3).But it neveroccurredto anyonein preexilictimes
to argue that such marriageswere null and void. Marriagewas the nonsacramental,private acquisitionof a woman by a man, and the state had
little or no legalstandingin the matter.5The foreignwomanwho marriedan
Israelitehusbandwas supposedto leave her gods in her father'shouse, but
even if she did not, it neveroccurredto anyone to arguethat her children
werenot Israelites.Sincethe idea of conversionto Judaismdid not yet exist
(see below),it neveroccurredto anyoneto demandthatthe foreignwoman
undergosome ritualto indicateher acceptanceof the religionof Israel.The
woman was joined to the house of Israel by beingjoined to her Israelite
husband;the act of marriagewas functionallyequivalentto the lateridea of
conversion.In somecircumstancesbiblicallaw and societydid pay attention
to maternalidentity-the childrenof concubinesand femaleslaves sometimes rank lower than the children of wives6-but it never occurredto
Sanders et al. (Philadelphia: Fortress Press, 1981), pp. 115-156, esp. 117-122. The abundant
literature spawned by the debate in the State of Israel on the question "Who is a Jew?" (see,
e.g., Baruch Litvin and Sidney Hoenig, eds., Jewish Identity: Modern Responsa and Opinions
[New York: Feldheim, 1965], a book brought to my attention by Rabbi Wolfe Kelman) is
important for the student of contemporary Jewry but not for the historian of antiquity. The
same judgment applies to the recent pronouncements of the Reform movement justifying a
patrilineal or nonlineal system (see the Winter 1983 issue of the Journal of Reform Judaism).
4. A convenient collection of the material is Raphael Patai, Sex and Family in the Bible
(New York: Doubleday, 1959), pp. 31-35. It was first collected in modern times by Leopold
Low, "Eherechtliche Studien," Gesammelte Schriften III, ed. Immanuel Lbw (Szegedin, 1893;
reprint ed., Hildesheim: Olms, 1979), pp. 112-143.
5. Most legal systems of antiquity shared this attitude. See, e.g., Elias Bickerman, "La
'
conception du mariage Athenes," Bulletino dell' Istituto di Diritto Romano 78, n.s. 17 (1975):
1-28.
6. See, e.g., Gen. 21:10, 22:20-24, and 25:5-6; Judg. 9:1-5 and 11:1-2; cf. Exod. 21:4.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
22
SHAYE J. D. COHEN
anyone to impose any legal or social disabilities on the children of foreign
women.
Although Deut. 7:3 prohibits Israelite men from marrying Canaanite
women, and Israelite women from being given in marriage to Canaanite
men, the legislator is particularly concerned about the former possibility,
because "their daughters will lust after their gods and will cause your sons to
lust after their gods" (Exod. 34:16, cf. Deut. 7:4).7This concern indicates not
a matrilineal principle but a patriarchal society in which the cult, as well as
everything else of importance, was the domain of the men. If Israelite men
are incited by their foreign wives to abandon the worship of the true God,
the result could be catastrophic; if Israelite women are turned astray by their
foreign husbands, who would notice?8 Furthermore, when an Israelite
woman was married to a foreigner she usually became part of her husband's
family (just like the foreign woman who was married to an Israelite
husband) and was no longer under the jurisdiction or authority of her native
people. Hence the legal and narrative texts pay little attention to marriages
between Israelite women and foreign men. Like the women themselves, they
were easily overlooked. A similar phenomenon can be observed not only in
the Mishnah (see below) but also in the literature of other societies which
enacted prohibitions of marriage with foreigners.9
The meager biblical data about such marriages suggest that the offspring
of Israelite women and foreign men were judged matrilineally only if the
marriage was matrilocal, that is, only if the foreign husband joined the
wife's domicile or clan.'0 1 Chron. 2:34-35 describes a marriage between an
There are many parallels in other legal systems of antiquity; see G. R. Driver and John Miles,
The Babylonian Laws, 2 vols. (Oxford: Clarendon Press, 1952), 1:350-353. A notable exception
is Egyptian law, according to which the offspring of a slave mother and a free father is legitimate (Diodorus of Sicily 1.80.3).
7. On Deut. 7:4 see below.
8. The Bible seldom notices idolatrous acts by Israelite women: see 2 Kings 23:7; Jer. 7:18
(the whole family); and Ezek. 8:14.
9. On Periclean Athens see below. Matroxenoi, "those born of foreign mothers," is an
attested Greek word (Pollux, Onomasticon 3.21 [E. Bethe ed., vol. 1, p. 161]; see the passages
listed in Bethe's note, especially the scholion to Euripides, Alcestis 989); patroxenoi, "those
born of foreign fathers," is unattested. The Digest gives the Roman version of the matrilineal
principle but omits the Lex Minicia (see below). (Mr. David Cherry suggests that it was omitted
because the Constitutio Antoninianaof 212 C.E.,which conferred Roman citizenship on virtually
all inhabitants of the empire, effectively rendered it obsolete.)
10. On matrilocal marriage in general, see Edward Westermarck, The History of Human
Marriage, 3 vols. (New York: Allerton, 1922), 1:296-297. On matrilocal marriage in the
ancient Near East, see Koschaker, "Fratriarchat, Hausgemeinschaft und Mutterrecht," pp.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
23
Egyptianslave and the daughterof his Israeliteowner. The marriagewas
probably a form of adoption." Other instances of matrilocalmarriage
involving a foreign male are recordedby Lev. 24:1012and 1 Chron. 2:17
(contrast2 Sam. 17:25).If the marriagewas not matrilocal,that is, if the
Israelitewomanjoined the house of her foreignhusband,I assumethat the
fellow nationalsof both the husbandand the wifewould haveconsideredthe
childrento be of the same nationalityas their father.Unfortunately,little
evidenceconfirmsthis assumption.'3
Ezra
The situation becomes less clear in the fifth century B.C.E.At about the
time that Pericles was enacting a law restricting Athenian citizenship to
those born of an Athenian woman lawfully wedded to an Athenian man
(451 B.C.E.),Ezra was promoting a similar reform in Judea.'4 After exacting
a pledge from the notables of Jerusalem to refrain from marriages with
foreigners, Ezra attempted to expel from the community approximately 113
foreign wives with their children. Only two aspects of this obscure episode
need to be treated here. First, why did Ezra virtually ignore the marriages
between Israelite women and foreign men?'5 Second, why did he attempt to
84-85; G. R. Driver and John C. Miles, The Assyrian Laws (Oxford: Clarendon Press, 1935),
pp. 134-142; E. Neufeld, Ancient Hebrew Marriage Laws (London and New York: Longmans,
Green, 1944), pp. 56-67; idem, The Hittite Laws (London: Luzac, 1951), pp. 126, 140-141, and
151-153; Werner Plautz, "Zur Frage des Mutterrechts im alten Testament," Zeitschriftfir die
alttestamentliche Wissenschaft74 (1962): 9-30, esp. 18-26; Patai, Sex and Family in the Bible,
pp. 52-53.
11. For the legitimation of the offspring of a slave father and a free mother, cf. Driver and
Miles, Babylonian Laws, 1:353-356.
12. The verse carefully contrasts the "son of an Israelite woman" with the "Israelite,"
thereby implying that the former was marked by a social stigma.
13. 1 Kings 7:13-14, cf. 2 Chron. 2:12-13.
14. On the citizenship law of Pericles, see W. K. Lacey, The Family in Classical Greece
(Ithaca, N.Y.: Cornell University Press, 1968), pp. 100-103; Jean-Marie Hannick, "Droit de
cite et mariages mixtes dans la Grece classique," L'antiquite classique 45 (1976): 133-148; John
K. Davies, "Athenian Citizenship: The Descent Group and the Alternatives," Classical Journal
73 (1977): 105-121; Cynthia Patterson, Pericles' Citizenship Law of 451-50 B.C. (New York:
Arno, 1981). The parallel between Pericles and Ezra was observed by George Foot Moore,
Judaism in the First Centuries, 3 vols. (Cambridge: Harvard University Press, 1927), 1:20, and
by Solomon Zeitlin, The Rise and Fall of the Judean State, 3 vols. (Philadelphia: Jewish Publication Society, 1968-78), 1:17.
15. He briefly mentions these marriages in his penitential prayer (Ezra 9:12) but otherwise
ignores them. So too his contemporary Malachi (2:11-16).
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
24
SHAYE J. D. COHEN
expel the children of the foreign wives as well as the wives themselves? The
usual answer to these questions is that Ezra introduced the matrilineal
principle. He attacked marriages between Israelite (at this period we can
begin to say "Jewish") men and foreign women because their consequences
were serious; like their mothers, the offspring were not Jewish. In contrast,
he could ignore (at least temporarily) the marriages between Jewish women
and foreign men because their consequences were relatively benign; like
their mothers, the offspring were Jewish.16 (Of course, according to the
Mishnah these offspring should have been expelled anyway, since they were
mamzerim-a minor detail.)
This view may be correct, but it is not necessarily so; other explanations
are possible. Perhaps Ezra ignored the marriages between native women and
foreign men because, as I have just mentioned, such marriages were generally ignored by both biblical and nonbiblical texts. Ezra's jurisdiction extended only to the members of his people,'7 and he could do nothing to a
foreign man who had married an Israelite woman. Even the attempted
expulsion of the children of the foreign wives does not necessarily presume a
matrilineal principle. Perhaps, like Pericles, Ezra introduced a bilateral
requirement for citizenship. This innovation affected matroxenoi (offspring
of foreign mothers) more than patroxenoi (offspring of foreign fathers) in
both Athens and Jerusalem, since the former had previously been regarded
as citizens while the latter had always been regarded as foreigners.'8 Furthermore, the attempted expulsion of the children was an act of supererogation by Shecaniah ben Jehiel (Ezra 10:2-3) and was not demanded by Ezra
himself (Ezra 10:11).19 If we insist on seeing the matrilineal principle in this
16. Chaim Tchernowitz, History of Hebrew Law [Hebrew], 4 vols. (New York: no publisher, 1943-44), 3:108-111; Samuel Belkin, "The Problem of Paul's Background," Journal of
Biblical Literature 54 (1935): 47-48; Zeitlin, Judean State, 1:27, referring to Solomon Zeitlin,
"The Offspring of Intermarriage,"Jewish QuarterlyReview 51 (1960): 135-140; Schiffman, "At
the Crossroads," p. 121. The amora R. Haggai also deduces the matrilineal principle from the
Ezra story (see below).
17. Ezra 7:25, reading ammak or ammehinstead of amma. See H. L. Ginsberg, The Israelian
Heritage of Judaism (New York: Jewish Theological Seminary, 1982), p. 17, n. 18.
18. Anastasios Christophilopoulos, Dikaion kai Historia (Athens: no publisher, 1973), p. 69
(part of his article "Marriage with a Foreign Woman According to Ancient Hellenic and
Hellenistic Law" [Greek]); C. Vatin, Recherches sur le mariage et la condition de la femme
maride l'depoquehellinistique (Paris, 1970), pp. 120-122; and the studies listed in note 14
above. Isaeus, Orations 8.43 summarizes the Periclean law as follows, "If our mother was not a
citizen, we are not citizens" (quoted by Lacey, Family in Classical Greece, p. 282, n. 14).
19. This was sensed by Thomas de Vio Caietan (1469-1534) in his commentary on Ezra
10:5, 11, and 44; see his Opera Omnia quotquotin Sacrae Scripturae Expositionem Reperiuntur,
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
25
story, we mustascribeits originnot to Ezrabut to an unheraldedmemberof
the clan of Elam.
The Second TemplePeriod
The likelihoodthat Ezra (or a contemporary)introducedthe idea that
the offspringof a Jewishfatherand a gentile motheris a gentile is further
diminishedby the fact that this half of the matrilinealprincipleis never
attested explicitly, and is frequentlycontradictedimplicitly,by the later
literatureof the Second Temple period.20When confrontingthe biblical
narrativesabout the heroes of Israel who married foreign women, an
exegete committed to the matrilinealprincipleand to the prohibitionof
intermarriagewith all gentileshas only threeoptions:(1) he can assertthat
the principleand the prohibitionwere in effect in biblical times, but that
they were occasionallyignored;(2) he can admit that the principleand the
prohibitionwere not in effectin biblicaltimes;or (3) he can arguethat the
principleand the prohibitionwere in effect and were observedby all. The
first two possibilitieswere near impossibilitiesfor anyone who believedin
the harmonyof SacredScriptureand Jewishlaw. Some rabbiscould admit
that Jacoband his sons did not observethe legal distinctionsbetweenclean
and unclean animals (M. Flullin 7:6),21 and some works of the pseudepi-
grapha could portray the patriarchs in exotic ways, but the general
5 vols. (Lyons: Jacobus et Petrus Prost, 1639), 2:370-371. Similarly, according to some Karaite
exegetes Ezra did not even attempt to expel the children; see below. Some modern Jewish
apologists explain that Ezra did not regard the women and their children as gentiles; he was
merely purifying the lineage of the aristocracy from admixture with Jews of poor pedigree. See
Heinrich Graetz, Geschichte der Juden, 2d ed. (Leipzig: O. Leiner, 1902), II, 2, pp. 130-136
(with the critique of Yehezkel Kaufmann, History of the Religion of Israel, vol. 4, From the
Babylonian Captivity to the End of Prophecy, trans. C. W. Efroymson [New York: Ktav, 1977],
pp. 340-342); F. Rosenthal, Monatschriftfiir die Geschichteund Wissenschaftdes Judentums30
(1881): 120-122; and Lewi Freund, "Iber Genealogien und Familienreinheit in biblischer und
talmudischer Zeit," in Festschrift Adolf Schwarz, ed. Victor Aptowitzer and Samuel Krauss
(Berlin and Vienna: R. L6wit, 1917), pp. 168-169.
20. Of course, one could argue that this fact is merely further testimony to Ezra's relative
lack of success; see, e.g., Morton Smith, Palestinian Parties and Politics That Shaped the Old
Testament (New York: Columbia University Press, 1971), pp. 121-122. The documents of
Elephantine describe several cases of intermarriage, but the matrilineal principle is nowhere in
evidence. See Bezalel Porten, Archives from Elephantine (Berkeley: University of California
Press, 1968), pp. 148-149, 203-213, and 252-258.
21. E. E. Urbach, The Sages, trans. I. Abrahams, 2 vols. (Jerusalem: Magnes Press, 1979),
1:335-336.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
26
SHAYE J. D. COHEN
tendencyof rabbinicliterature,Jubilees,and the Testamentsof the Twelve
Patriarchsis to claim that the patriarchsobservedthe law in a manner
consonantwith later beliefs.And even if an excusecould be found for the
marriagesof the sons of Jacob, what excuse could be found for Moses
himself and for the figureswho lived after the revelationof the Torah at
Mount Sinai? Rabbinic midrash, therefore,chose the third course, and
arguedthat the foreignwomen were not foreignat all, eitherbecausethey
were of Israeliteorigin or becausethey convertedto Judaismbefore their
marriages.22This exegesis does not necessarilypresume the matrilineal
principle-the removalof the blot of intermarriageis sufficientincentiveto
invent stories about Israelite lineage and conversions to Judaism-but
withoutit the matrilinealprinciplecannotbe maintainedin the presenceof
the biblical data. Therefore,the absence of this exegesis from the vast
majorityof the midrashicand exegeticalworksof the SecondTempleperiod
implies that these works are not familiarwith the matrilinealprinciple.
Some examples:The Book of Jubileeshas Abraham(20:4, 22:20, and
25:5), Rebecca(25:1-3), and Isaac(27:10)admonishtheirsons not to marry
Canaanitewomen, "for the seed of Canaanwill be rootedout of the land,"
but the author mentionswithout commentthe Canaanitewives of Simon
and Judah (34:20), marriageswhich were sanitizedby rabbinicexegesis.23
According to Jubilees 41:1 and Testament of Judah 10:1 Tamar, the
daughter-in-lawand paramourof Judah,was an Aramean;in rabbiniclore
(and in Philo) she was a convert to Judaism.24Testamentof Joseph 18:3
mentionswithoutcommentJoseph'smarriageto the daughterof the priest
of Heliopolis; rabbinicmidrash(and Joseph and Asenath) removes this
cause for opprobrium.25
EgyptianJews told manywonderfulstoriesabout
22. L. Ginzberg, Legends of the Jews, 7 vols, reprint ed. (Philadelphia: Jewish Publication
Society, 1967-68), index, s.v. "conversion" and "proselytes"; Victor Aptowitzer, "Asenath,
the Wife of Joseph," Hebrew Union College Annual 1(1924): 239-306; and Bernard Bamberger,
Proselytism in the TalmudicPeriod (1939; reprint ed., New York: Ktav, 1968), pp. 174-217. I
hope to study these techniques elsewhere.
23. R. H. Charles, Apocrypha and Pseudepigraphaof the Old Testament,vol. 2, Pseudepigrapha (Oxford: Clarendon Press, 1913), note ad loc.
24. In some midrashim, notably TargumJonathan on Gen. 38:6, she is the daughter of
Shem. (Her loyalty to God is indicated in TargumJonathan and Fragment Targum to Gen.
38:25.) See Marshall Johnson, The Purpose of the Biblical Genealogies (Cambridge: At the
University Press, 1969), pp. 156, 159-162, and 270-272 (who omits Ps.-Philo, Biblical Antiquities 9.5). For Philo (On the Virtues 40. 220-222) Tamar plays the same role as Asenath in
Joseph and Asenath: she is the perfect archetype of the female proselyte, just as Abraham is of
the male. See Marc Philonenko, Joseph et Aseneth (Leiden: Brill, 1968), p. 55.
25. See Aptowitzer, "Asenath."
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
27
Moses' Ethiopian wife, but one story they did not tell was her conversion to
Judaism; that motif appears only in the rabbinic version of the tale.26The
rabbis insist that a female war captive must convert to Judaism before being
married by her Israelite captor; Deut. 21:10-14 neglects this provision, as
does the Temple Scroll, a work closely related to Jubilees.27In the middle of
its paraphrase of Gen. 34, the story of Shechem's rape of Dinah, Jubilees
inserts a tirade against intermarriage. It condemns the unions of Israelite
men with foreign women and the unions of Israelite women with foreign
men, but, influenced by the scriptural text, directs the brunt of its anger
toward the latter, not the former. A man who gives his daughter to a gentile
is guilty of handing his seed over to Molech, and therefore both the man and
his daughter deserve death (30:7-11, esp.10). In rabbinic texts, by contrast,
the prohibition of Molech worship (Lev. 18:21) is understood to refer to
unions between Israelite men and foreign women: "this is a man who has
sexual intercourse with a gentile woman and raises up children who are the
enemies of God." Jubilees does not know the matrilineal principle, the
rabbis do.28
Following the biblical tradition, few texts of the Second Temple period
indicate the status of the offspring of unions between Jewish women and
gentile men. Some scholars have adduced Acts 16:1-3 to prove that such
offspring were considered to be Jewish, but the passage is ambiguous.
And he [Paul]came also to Derbeand to Lystra.A disciplewas there,named
Timothy,the son of a Jewishwomanwho was a believer;but his fatherwas a
Greek. He was well spoken of by his brethrenat Lystraand Iconium.Paul
wanted Timothyto accompanyhim; and he took him and circumcisedhim
becauseof the Jewsthat werein those places,for they all knewthat his father
was a Greek.29
Paul circumcises his new traveling companion so that he will not offend the
Jews of Asia Minor, but his act can be interpreted in either of two ways: is
26. Tessa Rajak, "Moses in Ethiopia: Legend and Literature,"Journal of Jewish Studies 29
(1978): 111-122, and Avigdor Shinan, "Moses and the Ethiopian Woman," in Studies in
Hebrew Narrative Art, Scripta Hierosolymitana, vol. 27 (Jerusalem: Magnes Press, 1978), pp.
66-78.
27. R. Akiva in Sifrei Deut. 213 (Finkelstein ed., p. 246); Temple Scroll, col. 63.
28. M. Megillah 4:9 and the statement of R. Ishmael in B. Megillah 25a; see S. Cohen,
"Prohibition of Intermarriage."
29. The translation is the Revised Standard Version.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
28
SHAYE J. D. COHEN
Paul circumcising an uncircumcised Jew, or is he converting a gentile to
Judaism? In other words, was Timothy a Jew with a gentile father or a
gentile with a Jewish mother? The latter possibility is supported by the
simple meaning of the phrase "for they all knew that his father was a
Greek." Patristic exegetes unfamiliar with the rabbinic matrilineal principle
had no doubt that Timothy was a gentile.30
Philo, Paul, and Josephus, three authors of the first century, are unfamiliar with either half of the matrilineal principle. Philo applies the term
nothos ("bastard") to the offspring of both types of mixed marriage: Israelite
mothers with gentile fathers, and Israelite fathers with gentile (or slave)
mothers. This usage is somewhat puzzling, but it certainly does not presume
familiarity with the matrilineal principle."3Paul thinks that either a Christian father or a Christian mother could "consecrate" the children to Christ
(1 Cor. 7:14). Josephus knows that priestly lineage can be debased if a
priest's wife is raped or is unfaithful,32but he does not know that a gentile
woman married to a Jewish man produces gentile children. Josephus carefully narrates the conversion (or, in some cases, nonconversion) of the
gentile men who married the princesses of the Herodian house, but he never
mentions the conversion (or nonconversion) of the gentile women who were
married to the princes of the house, and Josephus obviously has no doubt
that their children are Jewish. The biblical principle still applies: a gentile
30. S. Cohen, "Was Timothy Jewish?";cf. David Daube, Ancient Jewish Law (Leiden: Brill,
1981), pp. 25-26.
31. Israelite mother and gentile father: On the Life of Moses 2.36 ? 193 (referring to Lev.
24:10). Israelite father and gentile (or slave) mother: On the Virtues40 ? 224, and cf. Allegorical
Interpretation2.24 ? 94 (referring to the offspring of Bilhah and Zilpah); On the Life of Moses
1.27 ? 147 (referring to the mixed multitude, the offspring of Egyptian women and Hebrew
men). In some Greek cities the offspring of a citizen mother and a noncitizen father was a
nothos (see Demosthenes, 23 [Against Aristocrates] ? 213, referring to Oreus in Euboea), but I
think that Philo is using the term in a nontechnical sense: "of impure lineage" or "a product of
mixed breeding." This usage is well attested. Samuel Belkin argues that Philo knew the rabbinic
matrilineal principle, but the merits of his argument need not be considered (and they are few)
because he mistakenly believes that Philo restricts the term nothos to the offspring of Israelite
fathers and gentile mothers. See his Philo and the Oral Law (Cambridge: Harvard University
Press, 1940), pp. 233-235. The mistake originates with Isaac Heinemann, Philons griechische
undjiidische Bildung (Breslau: Marcus, 1932), pp. 313-314, but Heinemann, at least, realized
that Philo's usage is not compatible with rabbinic law. See too B. Cohen, "Law of Persons," p.
24, n. 71.
32. Jewish Antiquities 3.12.2 ? 276, 13.10.5 ? 292, 13.13.5 ? 372; cf. 11.3.10 ? 71 and Against
Apion 1.7 ? 35.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
29
woman "converts" by being married to a Jewish husband.33 Herod the
Great is labeled an "Idumean" and a "half-Jew" because of his paternal
ancestry; the fact that his mother was an "unconverted" Arab woman is
ignored. 4
Conclusion
The Apocrypha, the pseudepigrapha, the Qumran scrolls, Philo,
Paul, Josephus, and Acts are not familiar with the rabbinic matrilineal
principle. It is probably not a coincidence that most of these works also do
not knowthat foreignwomencan convertto Judaismby a meansotherthan
marriage to a Jewish husband. Conversion to Judaism by men is first
attested in the sixth and fifth centuries B.C.E.;its identification with circumcision is first attested (aside from the ambiguous story in Gen. 34) in the
second century B.C.E. Few texts of the Second Temple period, however,
speak of the conversion of women, and no text of the period prescribes
immersion as a component of the ritual of conversion for either man or
woman. (In Joseph and Asenath the ritual of Asenath's conversion is the
eating of honey from a honeycomb.) If the only way foreign women can
"convert" to Judaism is through marriage, the biblical material prevents
any matrilineal theory, since the Bible shows that the children of such
marriages were accepted as Israelites. With the emergence of the possibility
that women could convert to Judaism on their own, the matrilineal principle
could develop. This process is underway in the latter part of the Second
Temple period but is not complete until the second century, when immersion for women (and men) and the matrilineal principle are securely attested
for the first time.35
33. Jewish Antiquities 16.7.6 ? 225, 18.5.1 ? 109, 18.5.4 ?? 139 and 141, and 20.7.1-3 ??
139-147. (Josephus does not say whether Agrippa, the son of Drusilla and the uncircumcised
Felix [? 143], was Jewish.)
34. Herod: Jewish Antiquities 14.1.3 ??8-10, 14.7.3 ? 121, and 14.15.2 ? 403; cf. Eusebius,
Ecclesiastical History 1.7.11-14. Joseph Justus Scaliger noticed this problem and solved it by
arguing that Herod's mother converted to Judaism; see the Animadversionesin Chronologica
Eusebii, p. 149, in his Thesaurus Temporum:Eusebii Pamphili... ChronicorumCanonum ...
Libri Duo (Leiden: T. Basson, 1606).
35. For a brief history of conversion to Judaism, see S. Cohen, "Conversion to Judaism."
Several scholars have noted that conversion for women developed only after conversion for
men: see Tchernowitz, History of Hebrew Law, 1:301-303; Belkin, Philo, p. 235, n. 60; and the
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
SHAYE J. D. COHEN
30
The Mishnah
The central rabbinic text on the matrilineal principle is M. Kiddushin
3:12.36
[A] Whereverthereis potentialfor a valid marriageand the marriagewould
not be sinful,the offspringfollowsthe male.And what [woman]is this?Thisis
the daughterof a priest, Levite, or Israelitewho was marriedto a priest,
Levite, or Israelite.
[B]Whereverthereis potentialfor a validmarriagebut the marriagewouldbe
sinful,the offspringfollowsthe parentof lowerstatus.37And what [woman]is
this? This is a widow with a high priest,a divorceeor a "releasedwoman"
[balu;ah,Deut. 25:5-10] with a regularpriest,a mamzeretor a netinah[Ezra
2:43-58, etc.] with an Israelite,an Israelitewomanwith a mamzeror a natin.
[C]And any womanwho does not havethe potentialfor a validmarriagewith
this man but has the potentialfor a valid marriagewith other men, the offspringis a mamzer.And what [man]is this?This is he who has intercourse
with any of the relationsprohibitedby the Torah.
[D] And any woman who does not have the potentialfor a valid marriage
either with this man or with other men, the offspringis like her. And what
[man]is this? This is the offspringof a slave woman or a gentile woman.
Paragraph D clearly states that the offspring of a slave or gentile mother
(and a Jewish father) follows the mother, a ruling which appears also in M.
Yevamot2:5 and which constitutes half of the matrilineal principle.
This mishnah is carefully constructed out of four paragraphs which
share a common literary structure (each states a general rule followed by
one or more examples of the rule) but differ from each other in style.
Paragraphs A and B begin with kal makom she-yesh kiddushin,ask whether
chapter on conversion in Daube, Ancient Jewish Law. Elsewhere I hope to study the development of the rituals of conversion (including immersion) and the emergence of nonmarital
conversion for women.
36. I translate Albeck's text. For representative variants I have checked facsimiles of the
following codices: Parma De Rossi 138 and 984, Budapest Kaufmann A 50, Paris 328-329,
Leiden Scal. 3 (the most important manuscript of the Palestinian Talmud), and Munich 95 (the
most important manuscript of the Babylonian Talmud).
37. All the manuscripts listed in the previous footnote read (ha-)pagum she-bi-shneihen.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
OFTHEMATRILINEAL
PRINCIPLE
THEORIGINS
31
the unionis sinfulor not, ignorethe potentialof the womanto contracta
legal marriagewith other men, and introducewomen as their examples.
Paragraphs C and D begin with ve-kal mi she-ein lah alayv kiddushin, ask
whetherthe womanhas the potentialto contracta legalmarriagewith other
men, ignorethe sinfulnessof the union,and introducemen(morepreciselya
man in C and a masculinenoun in D) as theirexamples.In theirexamplesA
and C use a verb (she-niset/she-ba),B and D do not. These interlocking
patternsimplythat the examplesare an intrinsicpartof eachparagraphand
that the mishnahas a whole is an organiccomposition.38
The literaryunityof the mishnahis complementedby its thematicunity.
The mishnahis basedon two assumptionswhichare unknownto the Bible:
first, some marriagesare valid and some invalid;second, the statusof offspringis determinednot by the marriageof the parentsbut by the potential
of the parentsto contract a valid marriagewith each other (kiddushin).39
ParagraphA treats unions which are permittedand potentiallyvalid; B,
unions which are prohibitedbut potentiallyvalid; and C and D, unions
which have no potentialvaliditybecausethey are prohibited(althoughthis
point is left unstated).
The literaryand thematicunity of this mishnahimpliesthat each of the
four paragraphsis based on this postbiblicallegal theory,but the implication is misleading.The redactorhas broughtorderout of chaosby imposing
a single theory upon practices of disparateorigins. The basic principle
enunciated by A is biblical (the offspringof a union between Israelites
follows the father). C's ruling that incestuous relationships produce
mamzerimmay also be biblical,as some scholarshave argued.40Half of D
may be biblical(the offspringof a slave woman follows the motherat least
to someextent),41but the otherhalfcertainlyis not (the offspringof a gentile
woman follows the mother).Nor is thereany biblicalsupportfor the principle stated by B. A furtherindicationof the artificialityof the structureof
38. Albeck's edition uses ve-eizeh zeh to introduce the examples of A and B, and ve-eizo zo
to introduce the examples of C and D. Most of the manuscripts are not so exact.
39. "Potential to contract a valid marriage" is the only possible meaning of the term in
paragraphs C and D, a point not appreciated by many translators of this Mishnah. See B.
Cohen, "Law of Persons," p. 14. The rabbis do not require marriage of the parents for a child
to have the status of its father; see below.
40. The meaning of mamzer in Deut. 23:3 is obscure. See Neufeld, Ancient Hebrew
Marriage Laws, pp. 224-227; Encyclopaedia Biblica [Hebrew], vol. 5 (1968), s.v. mamzer;
Calum M. Carmichael, The Laws of Deuteronomy (Ithaca, N.Y.: Cornell University Press,
1974), pp. 173-174.
41. See note 6.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
32
SHAYE J. D. COHEN
this mishnah is the fact that two of the four examples in paragraph B are
inappropriate. Examples three and four (mamzer/mamzeret and natin/
netinah with an Israelite) well illustrate the principle that the offspring of
certain unions follow the parent with the lower status, but examples one and
two (widow with a high priest, divorcee and balurah with a priest) do not.
The children of these unions do not follow either parent; they are ballalim, a
status held by neither the father nor the mother.42
The Other Half of the Matrilineal Principle
Nor does M. Kiddushin3:12 account for all the anomalies of the rabbinic
law of status. The two Talmudim, the Tosefta (Kiddushin4:16 [Lieberman
ed., p. 292]), and the rabbinic commentators on this mishnah point out
many omissions, one of which is the union of a Jewish woman with a gentile
man. Like the Bible, the Mishnah is interested primarily in those who may
and may not be married by Israelite men. As the Palestinian Talmud says,
"Israelite men of good pedigree are prohibited from marriage with women
who are unfit, but Israelite women of good pedigree are not prohibited from
men who are unfit."43Elsewhere, however, the Mishnah does refer to this
half of the matrilineal principle. M. Yevamot7:5 states, without giving any
reason, that the child of a Jewish mother and a gentile or slave father is a
mamzer. Perhaps the ruling is a development of the principle stated by
paragraph D of M. Kiddushin3:12, which declares that the offspring of a
woman who does not have kiddushinfollows the status of the mother. The
logic of the argument is not spelled out, but most commentators assume that
the mother's incapacity to enter a valid marriage renders her child fatherless. Since there is no potential for de jure paternity, the child follows the
mother.44 Perhaps, then, this logic was extended also to men who lack
kiddushin.The offspring of a Jewish woman by a slave or a gentile does not
have a legal father and therefore follows the status of the mother. But just as
the law penalizes a Jewish man who has fathered a child on a gentile woman
42. In his commentary on this mishnah Maimonides tries to answer this difficulty.
43. P. Kiddushin3:14 64c; cf. B. Kiddushin73a and parallels. Neither the Babylonian nor the
Palestinian Talmud asks why the Mishnah omits the union of an Israelite woman with a gentile
or slave. The answer was obvious.
44. This is not, however, the logic of paragraph C; incestuous and various other prohibited
relations produce a mamzer, not a fatherless child who follows its mother.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
33
(the child is a gentile), the law had to penalize a Jewish woman who
conceived a child from a gentile man: the child is a Jew but a mamzer.
Roman law provides a good analogy to this process (see below). If this
exegesis is correct, paragraph D of M. Kiddushin3:12, as elaborated by M.
Yevamot7:5, testifies to both halves of the matrilineal principle.
The Talmudim, however, use not paragraph D but paragraph C for
determining the consequences of a union between a Jewish woman and a
gentile or a slave: since the mother does not have the capacity to contract a
valid marriage with the father (because he is a gentile) but does have that
capacity with other (i.e., Jewish) men, her offspring therefore is a mamzer.45
If this explanation is correct, one half of the matrilineal principle comes
from paragraph C, the other from paragraph D.
The two halves are even further apart if we accept the implications of T.
Kiddushin 4:16 (Lieberman ed., p. 292).
A gentileor a slavewho had intercoursewith an Israelitewomanandshegave
birth to a child-the offspringis a mamzer.R. Simeonb. Judahsays in the
name of R. Simeon, "There is no mamzerexcept [for the offspring]of a
womanwhose prohibitionis amongthe relationsprohibitedby Scriptureand
on account of whom [those who have intercoursewith her] are liable to
extirpation."
The first opinion, which is that of M. Yevamot7:5, declares that the offspring of a Jewish mother and a gentile father is a mamzer. R. Simeon,
however, says that the child is not a mamzer but a legitimate Jew.46This
debate, the Talmudim say, centers on the question whether or not paragraph
C ("the offspring is a mamzer") extends to the case under discussion. When
shorn of the talmudic additions, however, the debate in the Tosefta seems to
be based not on paragraph C of M. Kiddushin3:12 but on M. Yevamot4:13.
In that mishnah, which reflects neither the ideas nor the terminology of M.
Kiddushin 3:12, R. Akiva, Simeon the Temanite (perhaps the same R.
45. See P. Kiddushin3:14 64d and B. Yevamot45b (and parallels). The counter-argument in
both Talmudim is difficult to understand. That the principle of paragraph C includes male
gentiles and slaves seems to be confirmed by M. Gittin 9:2.
46. In medieval times some commentators (notably Rashi) argued that R. Simeon meant
that the offspring was not a mamzer but a gentile. See Aptowitzer, "Spuren des Matriarchats,"
p. 268. This view, incompatible with B. Yevamot44b-45b and contradicted by P. Yevamot4:15
6c (and P. Kiddushin3:14 64c-d), is a willful distortion of the Talmud and is a product of the
medieval period.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
34
SHAYE J. D. COHEN
Simeon quoted by the Tosefta), 47 and R. Joshua debate the definition of
mamzer. In their debate these three rabbis do not refer explicitly to the
consequences of a union between a Jewish woman and a gentile or a slave,
but T. Kiddushin4:16, by echoing the language of M. Yevamot4:13, implies
that this case too was included in their debate. If the Tosefta's interpretation
is correct, the Mishnah is unfamiliar with a single matrilineal principle. The
status of the offspring of a Jewish father and a gentile or slave mother was
determined by one legal principle (paragraph D of M. Kiddushin3:12), while
the status of the offspring of a Jewish mother and a gentile or slave father
was determined by another (M. Yevamot4:13 as understood by T. Kiddushin
4:16).
In sum, one mishnah clearly states that the offspring of a gentile mother
and a Jewish father is a gentile, and offers a reason for this ruling, although
the reason may not be the original one. Another mishnah clearly states that
the offspring of a Jewish mother and a gentile father is a mamzer, but offers
no reason for this ruling. The reason offered for the first ruling may be
applicable also to the second, although the Mishnah does not connect them.
Date
Both M. Kiddushin3:12 and M. Yevamot7:5 are anonymous. Can they
be dated? Jacob Neusner argues that M. Kiddushin3:12 is Yavnean (end of
the first century and beginning of the second century C.E.)because Rabbis
Eliezer and Tarfon refer to it in M. Kiddushin3:13, but this exegesis is incorrect. Rabbis Eliezer and Tarfon debate whether a mamzer can ever produce
children who are not mamzerim, but they do not discuss the definition of
mamzer and do not presume the principles outlined in 3:12.48 J. N. Epstein
argues that M. Kiddushin3:12 opens a tractate on "Forbidden Relations and
Pedigrees" which extends through 4:14, and which, aside from various interpolations, is an "ancient mishnah" of Second Temple times.49Unfortunately, this "tractate" is neither a thematic nor a literary unit, and the principles
47. The manuscripts and testimonia of T. Kiddushin 4:16 provide various readings; see
Lieberman's apparatus and commentary.
48. J. Neusner, A History of the Mishnaic Law of Women,5 vols. (Leiden: Brill, 1980), 5:173
and 200-201.
49. J. N. Epstein, Introduction to TannaiticLiterature, ed. E. Z. Melamed [Hebrew] (Jerusalem: Magnes Press, 1957), pp. 54 and 414-415. See Schiffman, "At the Crossroads," p. 118.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
35
of 3:12 nowhere recur within it. The discussion of the ten "genealogical
stocks" in chapter 4 includes many of the anomalous unions omitted by 3:12
(see above). But Neusner and J. N. Epstein, taken together, are correct to
the extent that M. Kiddushin 3:12-4:14 (or 4:12) is a series of anonymous
texts on genealogical matters, into which comments by named authorities,
almost all of them Yavnean, have been inserted. This implies that 3:12 is
Yavnean (if not earlier), but the conclusion is far from certain. If M.
Yevamot7:5 is to be connected with M. Yevamot4:13, as T. Kiddushin4:16
implies, it too is probably Yavnean.50
The two halves of the matrilineal principle are combined explicitly for
the first time in the Babylonian discussion of M. Kiddushin3:12. In order to
prove the scriptural origin of the ruling that the offspring of a gentile woman
follows the status of its mother, the two Talmudim quote (in slightly different forms) a statement of R. Yohlanan in the name of R. Simeon b. Yolhai.
The Babylonian version (B. Kiddushin 68b and parallels) appends the
following comment of Ravina, "Learn from this [the exegesis of R. Simeon]
that your daughter's son who is fathered by a gentile is called 'your son.' "
The Palestinian Talmud has nothing comparable to Ravina's comment,
perhaps because it thought that R. Simeon's exegesis extends also to the
offspring of Israelite mothers and gentile fathers (see below). If this is
correct, the two halves of the matrilineal principle were first connected by R.
Simeon (middle of the second century). According to the Babylonian
Talmud, however, the credit for this connection belongs not to R. Simeon
but to Ravina (late fourth century).
Dissenting Opinions
The anonymity of M. Kiddushin3:12 implies that the editor regarded its
rulings as beyond dispute. Virtually all tannaim and amoraim agreed that
the offspring of a gentile or slave woman follows the status of the mother.
Everyone further agreed that the offspring of a Jewish woman by a gentile or
a slave is a Jew, although there was a vigorous debate among the amoraim
whether to accept the Mishnah's view that such an offspring is a mamzer.
50. Neusner, Law of Women, 5:68 leaves M. Yevamot7:5 unassigned. A Yavnean date also
fits R. Ishmael's exegesis of Lev. 18:21 (see n. 28). Perhaps the matrilineal principle entered
rabbinic Judaism from the marriage regulations of pre-70 sectarianism (cf. Neusner, Law of
Women, 5:179-192), but I see no evidence to support this conjecture.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
36
SHAYE J. D. COHEN
The lenient view came to prevail, but in all other aspects the mishnaic statement of the matrilineal principle was followed.5'
The consensus was not unanimous, however. Rabbinic literature preserves several traces of an antimishnaic view (relics of the premishnaic
view?). Lev. 24:10 describes how a man, the son of an Israelite mother and
an Egyptian father, argued with an Israelite "in the camp," cursed God, and
was brought before Moses for judgment. The Sifra comments that the
phrase "in the camp" implies that the man had "converted" to Judaism
(Weiss ed., p. 104c). Nalhmanides and many other medieval commentators
are puzzled by this statement, for how can the son of an Israelite woman,
that is, a born Jew, convert to Judaism? None of the answers they offer is as
strong as the question. Whoever is responsible for this pericope of the Sifra
did not accept the dominant rabbinic view that the offspring of a Jewish
mother is a Jew.52 Similarly, some rabbis believed that the offspring of a
gentile mother and a Jewish father was a mamzer, not a gentile. R. Zadok in
Rome excused himself from spending the night with a woman who had been
offered to him because, he said, he feared that he might "multiply mamzerim
in Israel."53Even more extreme is the view of Jacob of Kefar Neburaya, who
argued that such a child was a Jew in all respects-the old patrilineal view of
51. On this debate see B. Cohen, "Law of Persons," pp. 15-19. The dynamics of this debate
remain to be investigated. The Talmudim ignore the evidence of 1 Chron. 2:34-35 (see above
note 11).
52. The reading of Weiss's edition of the Sifra is confirmed by codex Assemani 66, p. 465 in
the facsimile edited by Louis Finkelstein (New York: Jewish Theological Seminary, 1956). See
the following commentaries on Lev. 24:10: Nahmanides (Chavel ed., pp. 162-163), HIizzekuni
(Chavel ed., p. 410), Moshav Zekenim (S. Sassoon ed., p. 396), and Malbim. See Aptowitzer,
"Spuren des Matriarchats," pp. 274-277. Rashi on Lev. 24:10 had no difficulty with the Sifra,
since he thought that the offspring of a Jewish mother and a gentile father was a gentile; see
note 46. The end of T. Eduyyot refers to Lev. 24:10, but the text is corrupt.
53. Fathers According to Rabbi Nathan, version A, chap. 16 (Schechter ed., p. 32a).
Aptowitzer's interpretation ("Spuren des Matriarchats," pp. 266-267) is far-fetched: R. Zadok
feared that he might father a gentile son who might, in turn, marry a Jewish woman and father
mamzerim. See the ingenious suggestion of I. Heinemann, Zion 4 (1939): 284, n. 55 (mamzer is
here used as a translation of the Greek nothos) and the discussion of M. Herr, Studies in
Aggadah and Folk Literature, Scripta Hierosolymitana, vol. 22 (Jerusalem: Magnes Press,
1971), p. 137. Some medieval commentators found the same view in the statement of R. Asi in
B. Yevamot16b, "A gentile who betroths [a Jewish woman] at this time-we suspect that the
betrothal might be valid because he might be a descendant of the ten lost tribes." See the
Tosafot ad loc. with the comments of D. W. Halivni, Sources and Traditions:A Source-Critical
Commentaryon Seder Nashim [Hebrew] (Tel Aviv: Dvir, 1968), pp. 703-704, on B. Kiddushin
68a.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
37
the Bible! After a sternrebukefrom R. Haggai,he retractedhis opinion.54
All in all, these exceptionsdo not amount to much. Perhapsoutside the
rabbinicpale the matrilinealprinciplehad its opponents,but within rabbinic society it commandedalmost universalassent.55
III
The matrilinealprincipleis not attestedin eitherthe Bible or the literature of the Second Templeperiod. Even in the first centuryof our era it is
still unknownto Philo, the New Testament,and Josephus.All of thesetexts
seem to be familiarwith a patrilinealsystem.The Mishnah,however,states
the two halvesof the matrilinealprincipleas if theywereagreeduponby all.
Why did the rabbis break with previous practice?The innovation (or
innovations,if we keep the halves separate)can be explainedeither as the
productof a new ideology(a chapterin the historyof ideas)or as a response
to new social conditions (a chapter in social history). I shall treat each
approachseparately.
IdeologicalOrigins:Deuteronomy7:3-4
Both Talmudimusuallytry to find a basis in Scripturefor the rulingsof
the Mishnah,and both Talmudimadducethe exegesisof R. Yobananin the
54. P. Kiddushin3:14 64d; P. Yevamot2:6 4a; Genesis Rabbah 7:2 (Theodor-Albeck ed., pp.
51-52); and parallels. 0. Irsai suggests a different interpretation of Jacob's argument: Jacob
regarded the child as a convert to Judaism but believed that a child convert could be circumcised on the eighth day even if it was a Sabbath. See O. Irsai, "Ya'akov of Kefar Niburaia-A
Sage Turned Apostate" [Hebrew], Jerusalem Studies in Jewish Thought 2, no. 2 (1982/83):
153-168, esp. 157-163. (I am grateful to Dr. Marc Hirshman for this reference.)
55. The evidence for the practices of the nonrabbinic Jews of antiquity is meager. If gentiles
were allowed to erect epitaphs in the Jewish catacombs for their relatives who had converted to
Judaism, the sepulchral inscriptions of Rome provide no instance against the rabbinic matrilineal principle. See H. J. Leon, The Jews of Ancient Rome (Philadelphia: Jewish Publication
Society, 1960), appendix, inscriptions 68 and 222. Compare nos. 256 and 462; no. 21 is difficult
to understand, matrilineal principle or no. According to the Karaites the child of a gentile
mother and a Jewish father is a Jew. Therefore, they argued, Ezra expelled the foreign women
(because they had not converted to Judaism) but did not expel their children (since they, like
their fathers, were Jewish). See Bernard Revel, "The Karaite Halakah," Jewish Quarterly
Review 3 (1912-13): 375-376. Compare the latter-day apologists for Ezra (n. 19). (The
Karaites followed the rabbis in presuming that the Bible knew the institution of conversion for
women; see, e.g., Aaron b. Joseph the Karaite, Sefer ha-Mivbhar[Eupatoria, 1835], p. 17b,
commentary on Deut. 21:10-11 [the law of the female war captive; see above n. 27].)
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
38
SHAYE J. D. COHEN
name of R. Simeon b. Yobai to prove the scriptural origin of half of the
matrilineal principle. Here is the Palestinian version (P. Kiddushin3:14 64d,
cf. P. Yevamot2:6 4a):
It is written"Youshallnot intermarrywiththem:do not giveyourdaughterto
his son" [Deut. 7:3], and it is written"For he shall turnyour son away from
me" [Deut.7:4].Yourson froman Israelite[woman]is called"yourson,"but
your son from a gentile woman is not called "your son" but her son.
This version differs from the Babylonian (B. Kiddushin68b and parallels)
chiefly by having R. Simeon quote Deut. 7:3 as well as 7:4, but how the verse
or verses lead to the requisite conclusion is equally obscure in both
sources.56 In their commentaries on the Babylonian Talmud, Rashi and
Rabbenu Tam suggest that R. Simeon's deduction is based on the anomalous syntax of Deut. 7:4. Scripture prohibits the marriage of Israelite men
with Canaanite women as well as the marriage of Israelite women with
Canaanite men. Why, then, does Scripture say, "For he shall turn your son
away from me"? The verse should have read either "For she [or they] shall
turn your son away from me" or "For he shall turn your daughter away
from me." It is this anomaly which impelled R. Simeon to deduce that the
son of a gentile woman by an Israelite man is "her son" and not "your son."
So far Rashi and Rabbenu Tam agree; how the anomaly is connected with
the conclusion, however, is the subject of debate. Rashi argues that R.
Simeon understood Deut. 7:4 to refer to the marriage of an Israelite woman
with a gentile man: "for he [the gentile who married the Israelite woman]
shall turn your son [the son of his Israelite wife] away from me." R. Simeon
deduced that the son of an Israelite woman by a gentile man is "your son"
but that the son of a gentile woman by an Israelite man is not "your son."
Rashi's interpretation is perhaps corroborated by the Palestinian version,
which has R. Simeon quote "Do not give your daughter to his son" in addition to the first part of Deut. 7:3. This implies that R. Simeon thought that
the anomalous syntax of Deut. 7:4 referred to the marriage of an Israelite
woman with a gentile man, just as Rashi says.57
56. In the Babylonian version the anonymous redactor (the setam) quotes Deut. 7:3,
thereby splitting R. Simeon's exegesis in half.
57. The reading in P. Kiddushin3:14 64d is confirmed by codex Leiden Scal. 3. The parallel
in P. Yevamot2:6 4a (even in codex Leiden Scal. 3), however, omits the phrase "Do not give
your daughter to his son." Since scribes often exercised great freedom in copying biblical
quotations in the Talmud, my argument is, at best, suggestive.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
39
R. Simeonfindsa basisin Scripture
Accordingto Rashi'sinterpretation,
for both halves of the matrilinealprinciple.Accordingto RabbenuTam,
however, R. Simeon understoodDeut. 7:4 to referto the marriageof an
Israeliteman with a gentilewoman:"for he [thegentilefather-in-law]shall
turn your son [his son-in-law]away from me." Only half of the matrilineal
principleis derivedfrom Scripture.RabbenuTam'sinterpretationhas two
distinct advantagesover Rashi's:it is supportedby Ravina, who did not
think, as Rashi does, that R. Simeon treatedmarriagesbetweenIsraelite
women and gentile men (see above); and it is ratherclose to the simple
meaningof Deut. 7:4. For if we excludethe obviouspossibilitiesthat the text
of SacredScripturehas sufferedcorruption,or that the singular"he" and
"son" shouldbe understoodas equivalentto "they"and "sons,"the simple
meaningof the text is that of R. Simeonas explainedby RabbenuTam.58As
Exod. 34:16 demonstrates,Scripture is concerned primarily about the
marriageof Canaanitewomen with Israelitemen.
But even in RabbenuTam'sinterpretation,R. Simeon'sconclusioncan
be extractedfrom the scripturaltext only with great difficulty.The Babylonian Talmud admits as much by seeking scripturalproof-textsfor the
principlethat gentilesand slavesdo not havethe capacityto contracta valid
marriage,an exercisewhich would have been unnecessaryhad the proof
from Deut. 7:4 been convincing.59
The PalestinianTalmudadmitsas much
when R. Haggai, in order to silence the obstreperousJacob of Kefar
Neburya, deduces the matrilinealprinciplechiefly from Ezra 9-10, not
Deut. 7:4 (see above). PerhapsR. Simeonintendedhis exegesisto be merely
a scriptural hook from which to hang a rabbinic law (an asmakhta).
58. TargumNeofiti, TargumJonathan, TargumOnkelos, and Saadiah Gaon, followed by the
new (1962) Jewish Publication Society version, understand "he shall turn" to mean "they shall
turn" (Neofiti even supplies a subject, "their daughters shall turn"). Most of these also take
"son" to be the equivalent of "sons" ("children" in the JPS version). Rashi's exegesis is followed by R. Meyuhas b. Eliyahu, Commentary on the Book of Deuteronomy, ed. Yehiel Katz
[Hebrew] (Jerusalem: Mosad ha-Rav Kook, 1968), pp. 27-28; Rabbenu Tam's exegesis is
followed by R. Joseph Bekhor Shor in his Torah commentary, ed. Joseph Gad, 3 vols. (Jerusalem: no publisher, 1959), 3:16 (= Der Pentateuch Kommentar des Joseph Bechor-Schor zum
fiinfien Buche Moses, ed. Alfred Zweig [Breslau: Koebner, 1914; reprint ed., Jerusalem: Makor,
1978], p. 26), by Hjizzekuni(Chavel ed.), and by Aaron b. Joseph the Karaite, Sefer ha-Mivbar,
p. 6b (who discusses two other interpretations as well).
59. Medieval rabbis continued the exercise; see, e.g., the Sheiltot de Ray Abai Gaon, chap.
25 (ed. Samuel Mirsky, vol. 1 [Jerusalem, 1961], pp. 158-163), and B. M. Lewin, ed. Orar
ha-Ge'onim, vol. 9, Kiddushin (Jerusalem: Mosad ha-Rav Kook, 1940), p. 167, ? 376. Some
medieval rabbis were not convinced by any of these proof-texts; see the comment by R. Israel
in the Responsa of R. Asher b. Yebiel, ? 55.9 (Vilna, 1881, p. 52c).
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
40
SHAYE J. D. COHEN
Whateverhis intention,his statementdoes not disprovemy argumentthat
the matrilinealprincipleis unknownto Scripture.60
Ideological Origins:Ezra
If the matrilinealprincipledoes not come from Deut. 7:4, what are its
ideologicalorigins?In his refutationof Jacobof KefarNeburya,R. Haggai
anticipatesthe modernscholarswho suggestthat it was introducedby Ezra.
I demonstratedabove that this view is wrong, or, at least, not necessarily
correct,but perhapsR. Haggaiand his modernfollowersare correctto the
extent that the rabbisdeducedthe matrilinealprinciplefrom the Ezraepisode. Since the rabbis tried to avoid basing halakhahon any part of the
Bibleotherthanthe Pentateuch,we can understandtheirhesitationto admit
that the matrilinealprinciplewas derivedfrom the Book of Ezra. But the
difficultywith this suggestionis that accordingto the MishnahEzrashould
have expelled the childrenof the Jewish motherstoo. As mamzerimthey
could not remainas full-fledgedmembersof the community.6' The Ezra
episodecould have servedas a sourceonly for those who saidthat a Jewish
motherand a gentilefatherproducea legitimateJew, not a mamzer,a view
which did not gain ascendancyuntil amoraictimes.
Ideological Origins: The Uncertainty of Paternity and the Intimacy of
Motherhood
Scripture,then, does not providethe ideologicalorigins for the matrilinealprinciple.I offerthreeothersuggestions.First,perhapsthe principleis
based on the idea stated by Paulus in the Digest 2.4.5, semper [mater] certa
est, etiam si vulgo conceperit,pater vero is est quem nuptiaedemonstrant,"the
motheris alwaysknown,even if she conceivedout of wedlock,whereasthe
father is he whom the marriageindicates."The identity of a mother is
alwaysknowable,but the identityof a fatheris neverknowable;if a woman
60. Some modern Jewish apologists imply that the matrilineal principle really is to be found
in the sacred text. See S. D. Luzzatto, Commentaryto the Pentateuch [Hebrew] (1871; Tel Aviv:
Dvir, 1965), p. 520; David Hoffmann, Das Buch Deuteronomium,2 vols. (Berlin: Poppelauer,
1913-22), 1:91; and J. J. Hertz, The Pentateuch and Haftorahs (1936), 2d ed. (London: Soncino
Press, 1961), p. 775.
61. See note 19 above.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
41
is married,the law presumesthat her husbandis the fatherof herchild, but
this presumptionalways lacks certainty.Perhapsthe rabbis too believed
that paternitywas alwaysunknowableand felt that a child'sidentityshould
be determinedin the first instance by its mother and not by its putative
This view is unsatisfactoryfor two
father.Hencethe matrilinealprinciple.62
reasons. First, as I remarkedin the first paragraphof this essay, the rabbis
but paternityis
restrictedthe matrilinealprincipleto casesof intermarriage,
no more uncertainin those marriagesthan it is in unions betweenJews.
Paulustoo adducesthe uncertaintyof paternitynot in connectionwith the
Roman matrilinealprinciple(whichwill be discussedbelow) but in connection with somethingelse entirely(respectfor one's parents).Second,unlike
the Romans,the rabbisdid not requiremarriagebetweenthe fatherand the
motherfor the offspringto inheritfromhis fatheror receivehis status.If an
unmarriedwoman is pregnantand declaresthat the fatherof her child is a
priest, R. Gamalieland R. Eliezersay that she is to be believed;if a woman
becomespregnantas the resultof rape,the offspringis presumedto havethe
same status as the majorityof the people where the rape occurred(M.
Ketubbot1:9-10). In these cases paternityis very uncertain,but the rabbis
did not judge the offspringmatrilineally.63
Instead of emphasizingthe uncertaintyof paternity,some have suggested that the matrilinealprincipleis the result of the naturalcloseness
betweenmotherand child.64The offspringof a gentilemotherand a Jewish
fatheris a gentilebecausethe intimateconnectionbetweena motherand her
child makes it certain that she will influencehim and instructhim in the
waysof the gentiles.This suggestiontoo is unconvincing.The ancients,both
Jewish and gentile, recognizedthe intimacyof motherhood,but drew no
legal inferencesfrom this intimacy.Indeed,it was not until the nineteenth
centurythat the legal systemsof Europebeganto recognizethe legal rights
of a motherto her children.65
Accordingto rabbiniclaw a woman'sobliga62. Paulus' comment is usually summarized by the phrase mater certa, pater incertus. I have
not succeeded in finding a scholarly presentation of this view, but I believe that it is widely
accepted. At the 1983 meeting of the Society of Biblical Literature, it was advanced by several
people during the discussion of this paper. See below for another widely held view which lacks
scholarly documentation.
63. See Zeitlin, "Offspring of Intermarriage," p. 136. M. Ketubbot 1:9-10 is addressing the
status of the woman, not that of the child, but the implications are clear.
64. 1 have not found a scholarly presentation of this view, but it too was advanced during
the discussion at the 1983 meeting of the Society of Biblical Literature.
65. Priscilla Robertson, An Experience of Women: Pattern and Change in Nineteenth
Century Europe (Philadelphia: Temple University Press, 1982), p. 648, index, s.v. "child
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
42
SHAYE J. D. COHEN
tion to carefor herchildrenderivesmorefromherstatusas a wifethanfrom
her statusas a mother.It is the dutyof the fatherto raisethe children,andit
is the duty of the wife to assist him. The intimacyof motherhooddid not
have legal consequences.66
Ideological Origins: Roman Law
The citation from Paulusleads to my secondsuggestion:the matrilineal
principleenteredrabbinicthought underthe influenceof Roman law. The
parallelbetweenthe Roman and the rabbiniclaw of personswas observed
scholars,but it was Louis M.
by variouseighteenth-and nineteenth-century
Epsteinin 1942(followed, without acknowledgment,by Boaz Cohen)who
first suggestedthat the Roman law influencedthe rabbinic.67The suggestion, therefore,is not new; on what evidencedoes it rely?
Accordingto Roman law, a child is the legal heir of his fatherand is in
his father'scustody(potestas)only if his fatherand motherwerejoined in a
legal marriage (justummatrimonium).The capacity to contract a legal
and was possessed
marriagewas called conubium(also spelledconnubium),
almost exclusively by Roman citizens. Marriagebetween a person with
custody" (I owe this reference to Professor Paula Hyman). Intimacy of motherhood: Lacey,
Family in Classical Greece, p. 169 (quoting Xenophon, Memorabilia 2:2:5-10 and Aristotle,
Nicomachean Ethics 8:12:3 [1161B]). The rabbis know that a child often "honors" his mother
more than his father (B. Kiddushin 30b-31a and parallels).
66. Obligation of father to raise his children: M. Kiddushin4:14 and T. Kiddushin5:15-16
(Lieberman ed., pp. 297-298). Obligation of mother: T. Niddah 2:4-5 (Zuckermandel ed., p.
642) and M. Ketubbot 5:5. The intimacy of motherhood did have legal implications for the
treatment of animals: Sifra on Lev. 22:28 (Weiss ed., p. 99b) and B. Bekhorot 45. "The mother
is not of kin to her child," writes Henry Swinburne, A Treatise of Testamentsand Last Wills
(1591), quoted by Laurence Sterne, The Life and Opinionsof TristramShandy, vol. 4, chap. 29.
Some rabbis knew that a fetus was formed from matter contributed by both the mother and the
father (see B. Niddah 30b [and parallels] with the comments of J. Needham, A History of
Embryology, 2d ed. [Cambridge: At the University Press, 1959], pp. 77-79), but knowledge of
embryology had no impact on the development of law.
67. J. J. Wetstenius, Novum Testamentum Graecum, 2 vols. (Amsterdam: Officina
Dommeriana, 1752), 2:552 (commentary on Acts 16:3); Isaac Weil, Prosilytisme chez les Juifs
selon la Bible et la Talmud(Strassburg, 1880), pp. 79 ff., a work which I know only through the
citation by Bamberger, Proselytism in the Talmudic Period, pp. 89-90; M. Mielziner, The
Jewish Law of Marriage and Divorce (New York: Bloch, 1884), pp. 95-97; Epstein, Marriage
Laws, and B. Cohen, "Law of Persons." In his review of Epstein's book in Historia Judaica 6
(1944): 87-90, B. Cohen commented specifically on Epstein's suggestion (p. 89); why did he not
refer to it in his own essay on the subject?
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
43
conubiumand a personwithoutconubiumwas valid, but it was not ajustum
matrimonium; and without a justum matrimonium, the status of the child
follows that of its mother. Consequently,if a Roman citizen marriesa
noncitizen woman, the children are noncitizens.If a Roman citizen has
intercoursewith a slave woman, the childrenare slaves. Accordingto the
legal theory,if a Roman matronmarriesa noncitizen,the childrenare citizens, except that the Lex Minicia,a law probablyenactedduringthe first
centuryB.C.E.,declaredthat the childrenof such unions follow the parent
with the lower status, that is, the childrenfollow the father.Similarly,the
childrenof a Romanmatronby a slaveoughtto be, accordingto the theory,
free citizens like their mother, except that a law enacted under Claudius
declaredthat they are slaves.68
The differencesbetweenthe Roman and rabbinicsystemsare numerous
and important.For a son to follow his fatherthe Romansdemandnot only
conubium,the potentialfor a legal marriage,but alsojustummatrimonium,
the legal marriageitself. A child born of a matrimonium
injustumwas considered spurius or vulgo quaesitus. In M. Kiddushin3:12, however, the rabbis
use kiddushin to mean not "legal marriage" O'ustummatrimonium)but the
capacity to contract a legal marriage(conubium).Marriagebetween the
father and the mother is not essential,accordingto rabbiniclaw, for the
offspringto follow the father(see above).The rabbisdo not havea category
correspondingto the spurius,and the Romansdo not have a categorycorresponding to the mamzer.69In Roman law, even the offspringof an incestuous relationshipis nothingmorethan a spurius(Gaius,Institutes1.64and
Ulpian, Tituli5.7). A final distinction:the Romans allow their rulesto be
affectedby the intention of the actors. If a Roman citizen, either male or
female,marriesa noncitizenin the erroneousbelief that the noncitizenis a
citizen, and can demonstratethat the mistakewas an honest one, the non68. R. Taubenschlag, The Law of Greco-Roman Egypt, 2d ed. (Warsaw: Panistwowe
Wydawnictwo Naukowe, 1955), pp. 104-108; J. Crook, Law and Life of Rome (Ithaca, N.Y.:
Cornell University Press, 1967), pp. 36-68 ("The Law of Status"), esp. 40-41; A. Watson, The
Law of Persons in the Later Roman Republic (Oxford: Clarendon Press, 1967), pp. 27-28; M.
Kaser, Das r6mische Privatrecht,2d ed. (Munich: Beck, 1971), ??66 I 2, 68 I, 73 II, and 84 II;
Joseph Modrzejewski, "Un aspect du 'couple interdit' dans l'antiquit6: les mariages mixtes
dans I'6gypte hell6nistique," Le couple interdit: Entretiens sur le racisme, ed. Leon Poliakov
(Paris: Mouton, 1980), pp. 53-73, esp. 64-67 (a reference I owe to Professor Roger Bagnall).
The major primary texts are: Digesta 1.5; Ulpian, Tituli 5; Gaius, Institutiones 1.48-96. For the
law of Claudius, see Tacitus, Annals 12.53.
69. The closest rabbinic approximations to the spurius are the shetuki and asufi (M.
Kiddushin4:1-2).
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
44
SHAYE J. D. COHEN
citizen spouse and the children automaticallybecome citizens (Gaius,
Institutes1.67-68).70The rabbis have no parallelto this remarkablelaw,
althoughthey often accordintentionan importantplace in theirrulings.In
the law of status they make no such allowances.7'
These differences, however, cannot mask the conceptual similarity
betweenthe Roman and rabbinicsystems.Marriagesbetweencitizensand
marriagesbetween noncitizens72produce childrenwhose status is determined patrilineally.Marriagesbetween citizens and noncitizensproduce
childrenwhose status, in theory at least, is determinedmatrilineally;but
both legal systemstried to equalizethe consequencesfor male and female
citizens who strayed from the fold. A Roman matron impregnatedby a
noncitizen or a slave bears a noncitizenor slave, not a citizen; a Jewish
woman impregnatedby a gentile or a slave bears a mamzer,a citizen of
impairedstatus.
Even the languageof M. Kiddushin3:12 echoes Roman legal terminology. Ulpian, Tituli5.8 is a briefsummaryof the Romanlaw of status,and
its phraseologyis closely paralleledby M. Kiddushin3:12.73Conubiointerveniente liberi semper patrem sequuntur74 (compare the Mishnah's
"Wherever there is potential for a valid marriage . . . the offspring follows
the male") non intervenienteconubio matris condicioni accedunt75(compare
the Mishnah's"And any womanwho does not havethe potentialfor a valid
marriageeitherwith this man or with other men, the offspringis like her")
excepto eo qui ex peregrino et cive Romana peregrinus nascitur, quoniamlex
Minicia ex alterutro peregrino natum deterioris parentis condicionem sequi
70. See too Gnomon of the Idios Logos, ??39, 46, and 47.
71. The importance of intention in the Mishnah is analyzed and exaggerated by Jacob
Neusner, Judaism.:The Evidence of the Mishnah (Chicago: University of Chicago Press, 1981),
pp. 270-281.
72. See R. Yohanan in P. Yevamot2:6 4a and cf. B. Kiddushin67a; R. Hanina in B. Sanhedrin 57b. See the Roman sources analyzed by Kaser, Das rimische Privatrecht,? 73 II, and P.
E. Corbett, The Roman Law of Marriage (Oxford: Clarendon Press, 1930), pp. 96-106. The
offspring of a marriage between peregrini might follow the status of the mother if her nationality received a privilegium from the Romans to this effect; see Digesta 50.1.1.2.
73. "When conubium intervenes [between the father and the mother] the children always
follow the father; when conubiumdoes not intervene, they accrue to the status of the mother,
except that he who has a noncitizen father and a citizen mother is born a noncitizen, because
the Minician law ordains that one born of a noncitizen mother or father follows the status of
the lower parent."
74. Compare cum legitimae nuptiaefactae sint, patrem liberi sequuntur(Celsus in Digesta
1.5.19).
75. Compare ex eis inter quos non est conubium,qui nascitur . .. matris condicioni accedit
(Gaius, Institutes 1.78).
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
45
iubet'6 (comparethe Mishnah's"The offspringfollows the parentof lower
status").M. Kiddushin3:12is one of the few sectionsof the Mishnahwhich
can be translatedeasily into classicalLatin.7
Althoughit is generallyverydifficultto prove the influenceof one legal
The Roman law,
systemupon another,herethe evidenceis ratherstrong.7"
whose principlesare clearly attested in republicantimes," antedatesthe
earliestattestationof the rabbiniclaw. As far as I have been able to determine, if the rabbiniclaw had an externalsource,Romanlaw is the only real
possibility. The Greek law of the classical and Hellenisticperiods is not
comparable,nor is the law of Egypt or the other kingdomsof the ancient
Orient."0This suggestionaccounts for the phraseologyof the Mishnahas
well as its dominantideas. It also takesseriouslythe Mishnah'sexplanation
of itself, since the MishnahignoresScriptureand emphasizesthe notion of
kiddushin(a translationof conubium).It also is economical,sinceit accounts
at once for both halves of the matrilinealprinciple.
A full assessmentof this suggestionmust await a detailedstudyof other
possible influencesof Roman ideas and institutionsupon ancientJudaism.
If the matrilinealprinciplecan be shown to be but one of the manylegacies
of Rome to Jerusalem,the suggestionwill gain force. This study will also
have to addressseveral difficult questions. How did the rabbis learn the
principlesof the Romanlaw of status?Surelynot from the study of Roman
law books. How, then?Why did they allow themselvesto drawon Roman
76. Compare Gnomon of the Idios Logos, ? 39 (ta tekna hettoni genei akolouthei).
77. A point unappreciated by Guilielmus Surenhusius, Seder Naschim sive Legum Mischnicarum Liber qui Inscribitur de Re Uxoria (Amsterdam: Borstius, 1700), p. 377, who translates
the Mishnah literally (quocunque in loco obtinent desponsationes et non obtinet transgressio
infans sequitur virum, etc.).
78. For the methodological difficulties, see, e.g., R. Yaron, "Jewish Law and Other Legal
Systems of Antiquity," Journal of Semitic Studies 4 (1959): 308-331; Zeev Falk, "Zum fremden
Einfluss auf das juidische Recht," Revue internationale des droits de I'antiquite',3d series 18
(1971): 11-23; and Bernard Jackson, "On the Problem of the Roman Influence on the
Halakah," in Jewish and Christian Self-Definition, vol. 2, Aspects of Judaism in the GrecoRoman Period, ed. E. P. Sanders et al. (Philadelphia: Fortress Press, 1981), pp. 157-203 (with
extensive bibliography).
79. Cicero, De Natura Deorum 3.18 ? 45 and Topica 20. Livy 43.3.1-4 (a reference I owe to
Mr. David Cherry) seems to imply that the principle was already in force in 171 B.C.E.
80. The status of the offspring of mixed marriages varied from one Greek state to another;
see Vatin, Recherches sur le mariage, pp. 123-128 and Hannick, "Droit de cite et marriages
mixtes." In Roman Athens the offspring of an Athenian mother and a slave or noncitizen
father was a noncitizen; see Dio Chrysostom, Orations 15.3 (trans. T. Wiedemann, Greek and
Roman Slavery [Baltimore: Johns Hopkins University Press, 1981], p. 225). For the law of the
ancient Orient, see n. 10.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
SHAYE J. D. COHEN
46
wisdom when hatredof Rome was so widespreadin Judeansociety, even
among those opposedto war with Rome?These questionsare analogousto
those which must be asked in any study of rabbinic"Hellenism,"and the
answersremainelusive.
Ideological Origins: ForbiddenMixtures
My third suggestionsees the matrilinealprinciplenot as the result of
externalinfluencebut as an organicpart of rabbinicthought. The Babylonian Talmudasks whetherthere is scripturalsupportfor the notion that
slaves lack the capacityto contracta valid marriageand gives as its answer
Gen. 22:5, Abraham'sstatementto his servants"You stay here with the
ass."This is interpreted,in a veryunecumenicalpun, as "You stayhere,you
who are a race similar to an ass" (im
ha-domeh la-bamor).
ha-.amor/am
And since asses are unableto contractvalid
marriages,so too gentileslaves
Kiddushin
are unableto contractvalidmarriages(B.
68a). I am not aboutto
suggest that this oft-repeatedexegesis accuratelyportraysthe ideological
origins of our mishnah-I am not even sure of its real intents--but I am
suggestingthat the Mishnah'streatmentof the consequencesof intermarriageshould be juxtaposedto its discussionof the resultsof mixedbreeding
in the animal kingdom.82Scriptureprohibitsthe breedingof animals of
differentspecies (Lev. 19:9),but if the prohibitionis violated,what is the
statusof the resultingoffspring?Does it belongto the speciesof the fatheror
the speciesof the mother?Or is it a new speciesaltogether?In the Toseftathe
sages argue that a mule is neithera horse nor a donkey, but a new and
distinctspecies.It makesno differencewhetherthe mule'smotheris a horse
or a donkey;a mule is a mule (T. Kilayim5:5 [Liebermaned., p. 222]). The
81. Joseph Heinemann, "'A Race Similar to an Ass': The Transformation of a Midrash"
[Hebrew], Molad 22, nos. 193-194 (October 1964): 456-462. Heinemann observes that the
application of this exegesis to halakhic matters characterizes the Babylonian Talmud, not the
Palestinian. Heinemann argues that the original purpose of the exegesis was anti-Christian
polemic. More convincing is his argument that the exegesis should not be taken as rabbinic
"philosophy." Compare B. Hullin 5a (the sinners of Israel resemble animals) and Morton
Smith, "On the Shape of God and the Humanity of Gentiles," in Religions in Antiquity: Essays
in Memory of E. R. Goodenough, ed. J. Neusner (Leiden: Brill, 1968), pp. 315-326.
82. In his commentary on Deut. 7:4 R. Joseph Bekhor Shor observes, "[the offspring of] a
gentile or a slave follows the mother, just like an animal." For the structural parallel between
the classification of humans and the classification of animals, see Mary Douglas, Implicit
Meanings (London and Boston: Routledge & Kegan Paul, 1975), pp. 267-271.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
47
Mishnah, however, seems to ignore this opinion in favor of that of R. Judah
(M. Kilayim 8:4).83
R. Judahsays:All those bornfroma [female]horse,althoughtheirfatheris a
donkey,arepermittedone with another.Thustoo, those bornfroma [female]
donkey,althoughtheir fatheris a horse,are permittedone with another.But
those bornfroma [female]horsewiththose bornfroma [female]donkey,they
are prohibitedone with the other.
R. Judah's statement is ambiguous. Perhaps it means that a mule whose
mother is a horse and whose father is a donkey is permitted to mate not only
with other such mules but even with pure-bred horses, since a mule follows
the status of its mother.84If this interpretation is correct, M. Kilayim 8:4 and
M. Kiddushin 3:12 are exactly parallel. The offspring yielded by the mixed
breeding of either animals or humans is judged matrilineally: a mule, like the
offspring of a gentile mother and a Jewish father, belongs to the species of its
mother. The view of the sages (a mule is a separate species) has its analogue
in M. Yevamot7:5, since the status of a mule, which cannot mate with the
kind of either its father or its mother, is analogous to that of the mamzer, the
offspring of a Jewish mother and a gentile father (analogous but not
parallel-a mamzer is a Jew).85 It is most regrettable, therefore, that the
Babylonian Talmud rejects this interpretation and suggests that R. Judah
posits two types of mules, those whose mothers are horses and those whose
mothers are donkeys. Mules may mate only with their own kind (B. flullin
83. "Seems to ignore" because there is room for doubt; see M. Kilayim 1:6 with T. Kilayim
1:8 (Lieberman ed., p. 204). The rabbis knew that a mule is sterile, but the question had to be
treated anyway.
84. In their commentaries on the Mishnah R. Samson of Sens and R. Yom Tov Lipmann
Heller recognize that this is the simple meaning of R. Judah's statement. See too H. Albeck ad
loc.
85. B. Pesahim 54a explicitly compares the mamzer (the offspring of incest) to the mule:
Anah, the mamzer son of Esau, was the first man to mate a horse with a donkey. See TheodorAlbeck's discussion of Genesis Rabbah 82:14 in their edition, pp. 993-994. (I am grateful to Dr.
David Lieber for bringing this passage to my attention.) Cyrus the Great was a "mule" because
he was the son of a Persian father and a Median mother; see Herodotus 1.55.2 and J. Fontenrose, The Delphic Oracle (Berkeley: University of California Press, 1978), p. 302. C. M. Carmichael suggests that the deuteronomic prohibition of plowing with an ox and an ass together
(Deut. 22:10) is a veiled allusion to mixed marriage, especially the Dinah story of Gen. 34. See
his Laws of Deuteronomy, pp. 159-163. He expands upon this thesis in his "Forbidden
Mixtures," Vetus Testamentum32 (1982): 394-415, and Women, Law, and the Genesis Traditions (Edinburgh: At the University Press, 1979), pp. 33-48 and 57-73.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
48
SHAYE J. D. COHEN
78b-79a).A6According to this interpretation,the parallel between R.
Judah'sview and M. Kiddushin3:12is not exact. In the mixedbreedingof
humans, the result is either Jew or gentile, and the matrilinealprinciple
servesto determinewhetherthe offspringis the formeror the latter;in the
mixed breedingof animals,the resultis a hybrid,and the matrilinealprinciple serves to distinguishone sort of hybrid from another.
In any case, the importantpoint is that the laws of kilayim,prohibited
mixtures,providean ideologicalcontextfor the matrilinealprinciple.Jacob
Neusnerhas well demonstratedthe Mishnah'sdeep and abidingfascination
with mixturesand with creatureslike hermaphrodites,Samaritans,and the
land of Syriathat defy simpleclassification.87
For a reasonwhich I do not
this
did
fascination
not
extend
to
know,
gentiles who have become Jews
or
to
the
union of diverse kinds.
forbidden
(proselytes),
intermarriages,
Althoughthe Mishnahsays very little about these two categories,perhaps
what it does say, especiallyin M. Kiddushin3:12, should be understoodin
the light of its largerinterests.The offspringof mixedmarriageis a form of
kilayim.
Social Origins
I have surveyedfive possibilitiesfor the ideologicaloriginsof the matrilineal principle:Deut. 7:4, the Ezra episode, the uncertaintyof paternity
86. See too T. Kilayim 5:5 (Lieberman ed., p. 222). The discussion of Irving Mandelbaum,
A History of the Mishnaic Law of Agriculture: Kilayim (Chico, Calif.: Scholars Press, 1982) is
not entirely helpful. In his commentary on T. Kilayim 1:8, Mandelbaum writes (p. 47, n. 156,
referringto p. 324), "Cf. M[ishnah] 8:4, where Judah distinguishes between the mule (the dam
of which is a mare) and a hinny (the dam of which is a she-ass) and does not allow one to be
mated with the other. T[osefta] (which makes no such distinction) and M[ishnah] thus represent two different traditions of Judah concerning the same issue." On p. 270, however,
Mandelbaum explains R. Judah's opinion of T. Kilayim 5:5 in the light of M. Kilayim 8:4-so
the Tosefta is aware of the Mishnah's view. On p. 267, commenting on M. Kilayim 8:4, Mandelbaum writes, "According to Judah two mules which were born of dams of a single kind (and, of
course, of sires of a single kind as well) may be paired with one another, for these animals
themselves form a single 'kind.' " His parenthetic remark is far from obvious and needs greater
support than an introductory "of course." B. Hullin 78b adduces another instance of the
matrilineal principle in the animal kingdom: the prohibition of slaughtering an animal with its
young on the same day applies only to a mother and her young. See further Sifra on Lev. 22:28
(Weiss ed., p. 99b) and B. Bekhorot 45b.
87. Neusner, Judaism: The Evidence of the Mishnah, pp. 256-270. Neusner is applying to
the Mishnah the interpretation of Leviticus by Mary Douglas, Purity and Danger (London and
Boston: Routledge & Kegan Paul, 1966), pp. 52-53.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
49
theRomanlawof status,andthelawsof
(andtheintimacyof motherhood),
forbiddenmixtures.The last two are far more compellingthan the first
three. The major difficultywith all of them, however,is that they explain
how the rabbis could come to the matrilinealprinciplebut not why they
wouldwantto cometo it. Whatcompelledthemto departfromthebiblical
tradition and from the practiceof the Second Templeperiod?Did some
societal need dictate the new law?It has been suggestedthat many Jewish
women were raped by Roman soldiers during the wars of 66-70 and
132-135, and that the rabbis,out of pity for their plight, declaredthe resultingoffspringto be Jewish,not gentile.But the qualityof this suggestion
befits the obscurityof its origins,"8becausetelling an unfortunatewoman
who has been rapedthat she is aboutto beara mamzeris only slightlymore
consolatory than telling her that she is about to bear a gentile. In some
respectsit is less consolatory:a gentile,at least,can convertto Judaism,but
a mamzercan neverbe legitimated.89
Further,why declarethe offspringof a
Jewishfatherand a gentile motherto be gentile?If its purposewas to discourageintermarriageby Jewishmen, there seems to have been little need
for such legislation. Perhapsin first-centuryRome and AlexandriaintermarriagebetweenJewsand gentileswas not uncommon,but it certainlywas
uncommonin first-centuryJudeaand in rabbinicsocietygenerallythroughout the followingcenturies.90
And if the primarymotivationwas to restrain
rabbis
should
haveintroduceda bilateralrequirementfor
the
intermarriage,
Pericles
in
did Athens and Ezra did (perhaps)in Jerucitizenship,just as
salem (see above).
88. This view is widely held, especially in contemporary rabbinic circles, but I have not
found its original author; it is repeated by Daube, Ancient Jewish Law, pp. 27-30, but he too
does not indicate his source. The rape of women by victorious soldiers was a normal part of
ancient warfare (Isa. 13:16 and Zech. 14:2; see S. G. Cole, "Greek Sanctions against Sexual
Assault," Classical Philology 79 [1984]: 97-113, esp. 111-113, who refers to D. Schaps,
"Women of Greece in Wartime," Classical Philology 77 [1982]: 203-204), and it is likely therefore that many women were raped during the wars of 66-70 and 132-135 (see M. Ketubbot 2:9
with the Talmudim ad loc.; and Song of Songs Zuta, end, as presented by S. Lieberman, Greek
in Jewish Palestine [1942; reprint ed., New York: Feldheim, 1965], pp. 179-184). The question
remains whether their plight was the social setting for the emergence of the matrilineal
principle.
89. The amoraim reverse the ruling that the child is a mamzer, but the Mishnah is quite
clear on this point. See above.
90. S. Cohen, "Prohibition of Intermarriage."The problem with the lineage of the Jews of
Messene, etc., is not intermarriage but mamzerim and improper divorces; see B. Kiddushin
71b-72b (and parallels). Prof. Sarah B. Pomeroy impresses upon me that this suggestion
should not be rejected so lightly, since the purpose of Pericles' citizenship law (and its subsequent imitations) may have been to improve the marriageability of Athenian women (see
Lacey, Family in Classical Greece, pp. 100-103).
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
50
SHAYE J. D. COHEN
PrimitiveMatriarchy
In order to explain the matrilineal principle Victor Aptowitzer suggests
not societal need but social tradition. It is a relic of very ancient times
when Israelite kinship was matrilineal and Israelite society was matriarchal. Aptowitzer offers a striking parallel to the rabbinic law from
Herodotus' description of the Lycians (1.173):
Theircustomsare in partCretanand in partCarian.But they haveone which
is theirown and sharedby no othermen;theytaketheirnamesnot fromtheir
fathersbut fromtheirmothers;andwhenone is askedby his neighborwho he
is, he will say thathe is the son of sucha mother,andrecountthe motherof his
mother.Nay, if a womanof full rightsmarrya slave,her childrenaredeemed
pure-born;and if a true-bornLycianman takes a foreignwife or concubine,
the childrenare dishonored,though he be the first in the land.9'
This is the passage which launched J. J. Bachofen on his quest for Das
Mutterrecht, and this is the passage cited by Aptowitzer to prove that the
rabbinic law is a relic of primitive times.92Aptowitzer supports this thesis by
the discovery of numerous other "relics" in both the Bible and the Talmud,
none of them more striking than two mishnaic laws which appear on the
page following M. Kiddushin3:12. M. Kiddushin4:4 ordains that "He who
wishes to marry the daughter of a priest must investigate her lineage through
four mothers which are eight"-the pedigree of the bride's mother and
grandmothers is checked, but not the pedigree of her father and grandfathers. A proselyte, or the son of proselytes, may not marry the daughter of
a priest, and a priest may not marry the daughter of a proselyte. But if the
second-generation proselyte has a native-born Jewish mother, he or she
may marry into the priesthood (M. Kiddushin4:7). The same requirement
for a native-born Jewish mother appears elsewhere too.93
Is the matrilineal principle a relic of prehistoric times? Various scholars
91. The translation is that of A. D. Godley in the Loeb Classical Library, slightly modified.
92. J. J. Bachofen, Das Mutterrecht: Eine Untersuchungaber die Gynaikokratie der alten
Welt (Stuttgart, 1861). See Myth, Religion, and Mother Right.- Selected Writings of J. J.
Bachofen, trans. R. Manheim and introduced by J. Campbell (Princeton: Princeton University
Press, 1967). Aptowitzer, in "Spuren des Matriarchats," forgot to cite Bachofen.
93. M. Bikkurim 1:4-5, B. Yevamot45b, B. Kiddushin76b, and elsewhere. These passages
are misconstrued by Daube, Ancient Jewish Law, pp. 24-25. Compare Demosthenes 59
(Against Neaera), ? 106 (the Plateans who received Athenian citizenship may not become
archons, but their children born of native Athenian women may become archons).
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
51
have attemptedto findtracesof matrilinealorganizationin the legalsystems
of antiquity,but Aptowitzerdoes not aid his case when he confusesmatriliny (determinationof kinship through females)with matriarchy(rule by
females), a social form which never existed.94Whether ancient Israelite
society was ever matrilineal,I leave for othersto determine,but the alleged
relicsof that allegedsociety collectedby Aptowitzerare, for the most part,
trivialor debatable.95
I am not surehow M. Kiddushin4:4 and 4:7 shouldbe
but
relics
which
are nowhereattestedin the Bibleandpostbiblical
explained,
Jewishliteratureand which surfacemiraculouslyin rabbinictexts a millennium or two after the period of their origins-these are remarkablerelics
indeed. Perhapsa methodologicallysophisticatedstudy of rabbinicfamily
law and kinshippatternswill revealtracesof a matrilinealsociety,but in the
absence of such a study, Aptowitzer'ssuggestionis unconvincing.
IV
Summaryand Conclusions
In the biblicalperioda mixed marriagebetweenan Israeliteand a non94. E. Kornemann, "Mutterrecht," in Pauly, Wissowa, and Kroll, eds., Realencyclopadie
der klassischen Altertumswissenschaft,supplementary vol. 6 (1935), pp. 557-558; J. Bamberger,
"The Myth of Matriarchy," in Woman, Culture, and Society, ed. M. Z. Rosaldo and L.
Lamphere (Stanford: Stanford University Press, 1974), pp. 263-280; and the works listed by
Sarah Pomeroy, "Selected Bibliography on Women in Classical Antiquity," in Women in the
Ancient World. The Arethusa Papers, ed. J. Peradotto and J. P. Sullivan (Albany: State University of New York Press, 1983), pp. 347-354. For an erudite but disjointed discussion of
Herodotus 1.173, see Simon Pembroke, "Women in Charge: the Function of Alternatives in
Early Greek Tradition and the Ancient Idea of Matriarchy," Journal of the Warburg and
Courtauld Institutes 30 (1967): 1-35.
95. The rabbinic predilection for uncle-niece marriages is not a relic of a matrilineal
society, since most such societies would regard the marriage of a man with his sister's
daughter as incestuous. Rabbinic ideology and practice on this matter must be studied anew,
especially in light of the Qumran Temple Scroll, col. 66. See Samuel Krauss, "Die Ehe zwischen
Onkel und Nichte," in Studies in Jewish Literature in Honor of Kaufmann Kohler (Berlin: G.
Reimer, 1913), pp. 165-175; Louis Ginzberg, An Unknown Jewish Sect (New York: Jewish
Theological Seminary, 1976), pp. 23-24; Aptowitzer, "Spuren des Matriarchats," pp. 232-237;
and Salo Baron, Social and Religious History of the Jews: Ancient Times, 2d ed., 2 vols. (New
York: Columbia University Press, 1952), 1:310, n. 20, and 2:230 with 411, n. 14. Even matronymy is occasionally encountered in rabbinic sources (see S. Lieberman, Tosefet Rishonim, 4 vols.
[Jerusalem: Bamberger et Wahrmann, 1939], 3:160, and B. M. Lewin, Orar ha-Ge'onim, vol. 7
Yevamot [Jerusalem: no publisher, 1936], p. 156, ? 360; see too Bamberger, Proselytism, pp.
230-231), but in all likelihood this practice too is not a relic of primitive times. Compare A.
Christophilopoulos, "Matronymy among the Ancient Greeks" in idem, [Greek], Dikaion kai
Historia, pp. 60-67.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
52
SHAYE J. D. COHEN
Israeliteproducedoffspringwhichwas usuallyjudgedpatrilineally.
If an
Israelitewomanwasmarriedto a non-Israelite
man,shewouldjoinhisclan
and bearchildrenwho werenot Israelite.If he joinedher clan through
matrilocalmarriage,the childrenapparently
wereconsideredIsraelite.A
matrilocal
couldevenlegitimate
thechildrenof anIsraelitemother
marriage
and a slavefather(1 Chron.2:34-35).In the Mishnah,however,the childrenof an Israelitemotherand a gentilefather(eitherslaveor free)are
declarethesechildrento
Jewsof impairedstatus.TheTalmudim
mamzerim,
but full andlegitimateJews.Bothdecisions,at leastin
be not mamzerim
In biblicaltimesmany
casesof patrilocalmarriage,contradictthe Bible.96
Israelitemenmarriedforeignwomen,andtherewasneveranydoubtthat
thechildrenwereIsraelite.Theoffspringof a slavemotherandan Israelite
sufferfromsomedisabilities,
butno onequestioned
fatherdid,apparently,
its Israelitestatus.The Mishnah,however,explicitlystatesthat suchoffspringfollowthe mother,andthisrulingis not disputedin theTalmudim.
As faras I havebeenableto determine,
thetransitionfrombiblicalpatriliny
cannot
be
dated
beforethe periodof the Mishnah
rnishnaic
to
matriliny
to introducethematrilineal
itself.Thereis no evidencethatEzraattempted
principle,and evenif he did, thereis abundantevidencethat it was still
unknownin thefirstcenturyof ourera.WhydidtheYavneanrabbisdepart
frombiblicalpractice?
Therearetwo goodexplanations,
one internal,the otherexternal.The
the mishnaiclaws regardingthe
accords
with
matrilineal
principle
nicely
mixtureof diversekinds(kilayim).Theunionof a Jewwitha gentileis akin
to theforbiddenunionof a horsewitha donkey.In bothcasestheMishnah
Evenmorestrikingis theparaljudgestheresultingoffspringmatrilineally.
lel affordedby the Romanlaw of status.The terminology,ideas,and
3:12arethoroughly
Roman:if oneparentdoes
of M.Kiddushin
conclusions
in Latin,
not possessthe capacityto contracta legalmarriage(conubium
in Hebrew),theoffspringfollowsthemother.Therabbis,likethe
kiddushin
Romans,departedfromthisprinciplein orderto penalizea citizenwoman
thattheoffspring
whomarrieda noncitizenor a slave:theRomansdeclared
followsthe parentof lowerstatus(in this case, the father),the rabbis
I amunableto decidebetweenthese
declaredthattheoffspringis a mamzer.
two explanations.Perhapsthey are both true, the rabbinicnotionsof
kilayimfacilitatingthe influxof the Romanlaw.
96. Why the Talmudim reversed the Mishnah, and why some medieval scholars reversed
the Talmud (n. 46), requires investigation.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions
THE ORIGINS OF THE MATRILINEAL PRINCIPLE
53
Another factor is relevanttoo. The idea of conversionto Judaismis a
creationof the exilicperiod.At firstit was an option only for men;its ritual
was circumcision. A gentile woman "converted" to Judaism through
marriagewith a Jewish husband,a procedurepresumedby the Bible and
still presumedby Josephus.Gradually,however,conversionfor womenwas
introduced;its ritualwas immersion(a practicewhich also becamepart of
the conversionritualfor men). This facilitatedthe rise of half of the matrilinealprinciple,sincethe gentilewomanwas now a personwhoseJewishness
could be determinedwithout referenceto her Jewishhusband.If she converts to Judaism,the childrenshe bears to her husbandare Jewish;if she
does not, they are gentile, in spite of the Jewishnessof her husband.This
new ideology mandatedthe reinterpretationof the biblicalnarrativeswhich
glibly admittedthat the heroes of ancient Israel marriedforeign women.
Thesedevelopmentsare obscure,but they certainlyformpartof the ideological backgroundto the emergenceof the matrilinealprinciple.97
All of these suggestionsare exercisesin intellectualhistory. Did social
historytoo play a rolein the creationof the matrilinealprinciple?Numerous
practitionersof Jewish Wissenschafthave argued that rabbinic law was
determined,at least in part, by the social and economic needs of contemporary Jewry. The matrilinealprinciplehas had enormous social consequencesfor modernJews,and it is easy to believethat the rabbismusthave
been compelled by some societal need to institute it. But there is little
evidenceto supportthis belief. Intermarriagewas not a severeproblemin
rabbinicsociety,and even if it was,the logicalresponsewouldhavebeenthe
institutionof a bilateralsystem(eithera gentile fatheror a gentile mother
rendersthe offspringa gentile).Perhapsin regardto othermattersthe rabbis
werelegislatorslisteningattentivelyto the demandsof theirconstituency.In
their statementof the matrilinealprinciple,however,the rabbiswerephilosophers, and, like most philosophers,they did not always live in the real
world.
JewishTheological
of
Seminary
America
New York,N.Y.
97. Perhaps this development is connected with a change in the status of the woman in
ancient Judaism.
This content downloaded from 128.103.149.52 on Mon, 19 Jan 2015 11:38:23 AM
All use subject to JSTOR Terms and Conditions